Re: Re: Old CNC controls

"D.F.S." wrote:

> From: "D.F.S." <dfs@...>
>
> >
> > From: Jon Elson <jmelson@...>
> >
> > "D.F.S." wrote:
> >
> > The threading does not use some fixed value for spindle speed, it
> > measures
> > spindle rotation with a spindle encoder, and each sampling interval
> it
> > computes
> > a new carriage position based on the spindle position.
>
> Is this something EMC does "Out of the Box"?
>
> I figured this would be the Best approach, but thought EMC could only
> pre-calculate the positions of all the axis at any given instant, and
> compensate to try and force everyting back into sync.
> The only "Correction" is if things fall far enough out of sync, past
> a defined parameter, was to simply stop with an error.
>
> This is not the way EMC works?

No, there is a G-code for synchronized threading. Apparently, there are
two
very similar codes, one for mills (G84) which is a tapping cycle, which
reverses
the spindle at the end of the move, and backs the tap out of the hole.
This has
apparently been implemented, although it is not in the interpreter
documantation.
I was assured by Fred Proctor (or was it Matt Shaver?) about a year ago
that
GM Powertrain had tested it on one of their big machining centers, and
it worked.

The turning center G-code is G33, for constant lead threading, and it
doesn't
require you to back up the spindle to return the tool to the start end
of the thread
to take another pass. G34 does variable lead threading, which sounds
like an
obscure thing to do.

I don't believe this is implemented, but if G84 has been done, G33
should be no
problem to add.

> Where is this compensation feature documented or is it a DIY kind of
> function?

Well, it is not a 'compensation'. You need to install a spindle
encoder, with an
index channel, and hook it up to an encoder input on the STG board (or
whatever
you are using). EMC computes where the Z axis should be relative to
spindle
position.

> I'll have to look into a control for my motor.
> What is the usual speed variation that can be acheived had with one of
>
> these controls?
> I would presume there is only a limited frequency change that will
> properly operate a motor designed for a fixed frequency.

Not really. Cooling is the major problem, as most AC motors use
self-driven cooling
fans. If you provide your own, constant-speed cooling fan, then the
motor can be
run down to nearly zero RPM. With external cooling, it should be no
problem at
all to run from 15% to 150 % or so of rated speed. You get constant
torque up
to 100% rated RPM, then constant power above that.

>
> > I built my own PWM servo amps, since I didn't know where to get them
>
> > surplus at the time. It was a great education in control systems
> theory
>
> I have a lot to learn, maybe it's time to pick my lessons.
>
> I'll build one if necessary, or if the cost appears too high to me,
> but
> I presume from that statement you NOW know of a source for surplus
> servo amps?

Well, they do come on the surplus market from time to time, not
necessarily
cheap, often hard to get documentation, etc. But, I've seen them on
eBay
and such places. I have made my design available to other people (about

20 individuals, so far) as a semi-kit. See my web page at :
http://ascc.artsci.wustl.edu/~jmelson/servo.html
for more info.

Jon